Routing Protocols

Routing comes from the term “route” which means the path to transport data packets in
computer networks. Routing is a process of moving data from one network to another
choosing the best path for transport. Routing Protocols are set of rules and algorithms
used to select the best path for transport for example, BGP, RIP, etc.

Types of Routing Protocol:

 Static:
 Non-adaptive approach
 Manual Technique where the network administrator selects the best path.
 Only admin has access to perform this action.
 Default:
 Router’s routing table:
o Router receives the packet, checks its routing table to find matching route
for the destination IP it received.
o If not found, it uses the pre-configured default route.
 Forwarding the Packet:
o The router forwards the packet to the next-hop router based on the
default route.
o Next-hop router also checks its routing table, if not found; forwards the
packet to the next-hop
 One hop to other until error:
o The packet is hopped from one to another router until it reaches the
destination. If a router doesn’t contain “next-hop” information, sends back
an error, “ICMP Destination Unreachable” to the sender.
  Dynamic:
 Work Process:
o Relies on routing protocols to discover best paths to the
destination.
o In case of error or better path’s availability, routers can share
information.
o Once change is detected, router updates its routing table with new
route information.
 Adaptive Routing

Types of Dynamic Routing Protocols:

1. Interior Gateway Protocol (IGP):

 Used within the same network or Autonomous Systems (AS)
 Main Function:
o Searches for the best route.
o Forwards data packets from sender to receiver within the same AS.
 Type of IGP:
o Link State Protocol
o Distance Vector Protocol
o Hybrid Protocol
a. Link State Protocol:
 Each router calculates the shortest path using Dijkstra’s algorithm.
 Exchange information about the status of their connection instead of number of
hops
 Converges faster than Distance-Vector Protocol.
 Examples: OSPF, IS-IS.

i. OSPF - Link-State Routing Protocol:

 Uses Dijkstra’s Algorithm for calculations.
 To determine the best route, it uses link cost (bandwidth, delay and reliability.
 Divides network into Areas to improve scalability, reduce overhead and
optimize route calculations.

 Working Process:
 Neighbor Discovery:
o Sends a “Hello” packet to directly connected routers.
o Forms R-R relationship with adjacent routers to exchange information.
 LSA Exchange:
o Each router gathers info about directly connected links (bandwidth,
status, and cost). The routers exchange Link-State Advertisements
(LSAs).
o LSAs contain link cost and topology information, which routers use to
calculate the best path.
o Only changes in the LSA is shared, not the entire database.
o Stored in the common Link-State Database where, each router have
a copy and is synchronized across all routers.
 Route Calculation:
o Each router runs Dijkstra’s Algorithm to calculate best path to every
destination.
o Best path is stored in the OSPF routing table.
 Packet Forwarding:
o After the completion of the routing table, data packets are forwarded
through the most efficient path.
o If network changes occur, only the affected routers send updates,
making the overall process efficient.

 Areas:
 Backbone Area (Area 0):
o Main Highway
o Responsible for interconnecting other areas.
o All non-backbone areas must connect to this area for proper routing.
 Regular or Standard Area:
o Contains internal routers and networks that exchange routing
information.
o Receives summarized routes from other areas via Area 0.
ii. IS-IS - Intermediate System to Intermediate System:
 Primarily used and preferred by larger enterprises and ISPs.
 Uses Dijkstra’s Algorithm.
 Doesn’t require backbone Area 0.

 Working Process:
 Neighbor Discovery:
o Routers exchange “Hello” packets and form adjacencies when
settings match.
 LSP Exchange:
o Similar to LSA but the structure is different.
o Each router creates a Link-State PDU (LSP) containing link
information.
o LSPs are flooded to all routers within the same area (Level 1) or
between areas (Level 2)
 LSDB Synchronization:
o Same as OSPF, a common synchronized database maintained
across routers.
 Path Calculations:
o Best routes are stored in the Forwarding Table (FIB).
 Packet Forwarding:
o If network change occurs, IS-IS updates only the affected LSPs.

 IS-IS Levels (Level 1 & 2):

 Level 1 Routers (L1) - Local pathways in a neighborhood.
o Communicates only within the same area.
o Oblivious of other areas.
o Forwards unknown destinations to L1/L2 Router.
 Level 2 Routers (L2) - Highways connecting cities
o Handles inter-area communications
o Can directly route between areas without needing a backbone.
 L1/L2 Routers (L1/L2) - Like city exits that connect local road to the
highways
o Combination of L1/L2.
o Can route inside an area and between areas.
o Help L1 routers forward traffic to other areas.
b. Distance-Vector Routing Protocol:
 Introduction:
 Distance = Measurement of paths & Vector = direction to which the
packets are forwarded to.
 Finds best path using, Bellman-Ford Algorithm: (For more info on how
Bellman-Ford works, tap the link.):
o Determined best path via metrics, hop count.
 Sends routing table periodically to neighbors.
 Prone to routing loop.
 Looping Control Mechanism implemented:
o Split horizon:
 Prevents a router from advertising a route back to its
source of information (its neighbor).
o Route Poisoning or Poison Reverse:
 When a route becomes unreachable, it is advertised with
metric of infinity (16 in RIP) to force removal.
o Hold-Down Timers:
 When a router detects a route failure (e.g. a link goes
down).
 Timeout is enforced onto the router after it receives an
update. Such that any information passed immediately
after that cannot reach the router.

 Process Involved:
 Initialization:
o Each router starts by discovering directly connected networks
o Creates an initial routing table using:
 Directly connected network
 Next-hop router
 Cost for each route
 Sharing Updates:
o Sends entire routing table to directly connected neighbors at
regular intervals.
 Update Routing Table:
o Updates are propagated hop-to-hop.
o When a router receives an update:
 New route- It adds it to the routing table
 Better route - Updates existing routing table
 Same route - Ignore
o If Link fails:
 Marks the route as unreachable (metric = infinity)
 Update is sent to all neighboring routers
 Loop prevention mechanism activates.
i. Routing Implementation Protocol (RIP):

 Introduction:
 Determines best path using hop count.
 Maximum hop limit = 15.
 Sends routing table every 30 seconds.

 Working Process:
 Router builds its routing table with directly connected devices
 Every 30 seconds, router broadcasts its routing table to neighbors
 RIP chooses route with lowest hop count.
o If multiple route has same hop count, it uses equal-cost load
balancing.
 Updates are shared hop-by-hop.
 If router detects an error:
o Sends “Unreachable” = if hop count = 16
o Uses UDP port 520 for communication.

 Versions:
 RIP v1:
o Original version of RIP
o Viable for small networks
o Doesn’t support CIDR (Classless Inter-Domain Routing):
 Only works with Class A, B or C networks
o Sends update to all routers causing increase in traffic.
o This doesn’t support authentication, vulnerable to spoofing attacks.
 RIP v2:
o Supports both CIDR and variable-length subnet mask (VLSM).
o Only broadcasts to RIP enabled routers.
o Uses MD5 authentication to avoid fake injection of routes to its routing
table.
 ii. Interior Gateway Routing Protocol (IGRP):

 Introduction:
 Best path found using composite metric: Bandwidth, Delay, Load and
Reliability.
 Max hop count = 255.
 Updates after every 90 seconds.
 Uses UDP port 9 for communication.
 Unsupported in non-cisco devices.

 Working Process:
 Builds an initial routing table.
 Updates after every 90 seconds to its neighbors.
 If better metric is discovered, neighbors update their routing table.
 Best route is selected.
 If route fails, IGRP marks it as “Unreachable”.

c. Hybrid Routing Protocol:

 Introduction:
 combination of Distance-Vector and Link-State
 Uses Distance-Vector principles:
o Shares routing information with neighbors like RIP or IGRP
 Uses Link-State features:
o Maintains a topology table and updates only necessary changes, like
OSPF.
 Updates only affected route to avoid loops
 hybrid protocols support sub-netting and super-netting.
 Loop Prevention:
o Uses Feasible Successor routes (EIGRP) or AS-Path Filtering (BGP).
 Faster Convergence:
o Recalculates routes using advanced algorithms like DUAL (Diffusing
Update Algorithm) in EIGRP.
i. Enhanced Interior Gateway Routing Protocol (EIGRP):

 Introduction:
 Improvement over IGRP.
 fast convergence, low bandwidth consumption, and advanced metric
calculations.
 Uses DUAL Algorithm.
 Supports Classless Routing (CIDR & VLSM).
 Allows backup route in case of route failure.
 Supports both IPv4 and IPv6.

 Working Process:
 Step 1:
o Discover neighbors by sending Hello Packets.
o Routers store neighbor’s details in the Neighbor Table.
 Step 2:
o Exchange routing information via Update Packets.
o Routers store data in Topology Table.
 Step 3:
o Uses DUAL Algorithm to find the best path.
o The route with the lowest feasible distance (FD) is selected as the
Successor.
o If an alternative path exists with a feasible condition (reported
distance < FD), it is stored as a Feasible Successor.
 Step 4:
o Best route is installed into the routing table.
o If successor fails, switches to the backup path.
o If no Feasible Successor is available, EIGRP enters Active Mode
and queries neighbors for a new path.
 Step 5:
o Continuously monitors the network using periodic “Hello Packets”.
o If a route fails, the router checks the Feasible Successor List for an
alternate path.
o If no alternate path exists, the router queries its neighbors and waits
for replies.
2. Exterior Gateway Protocol (EGP):

 Introduction:
 Exchange routing information between different Autonomous Systems (AS)
in the internet.
 Used by ISPs and large networks to exchange routes.
 Based on Path-Vector Routing.
 Maintains AS Path information to prevent loops.
 Type of EGP: BGP or Border Gateway Protocol

 Working Process:
 Two routers in different AS establish a connection.
o They exchange hello messages.
 Routers periodically send “Keep Alive” messages.
o If no response is received, the router is deemed unreachable.
 Protocol selects the best path on the basis of policies.

a. Border Gateway Protocol (BGP):

 My Logic:
Following after the logic of NAT where, modified package is received by the ISP.
The ISP uses routing protocol to direct or route the package to its destination.
BGP is used to find the best route for data transfer from one network to another.
It does this by enabling connection between two Autonomous Systems (AS)
where, AS is the collection of networks (routers and their connected devices)
under control of a single organization or administrative authority which follows a
unified routing policy.

 Key Points:
 Inter-network connection
 Finds best route for data share
 Can re-route in case one path fails
 Runs over TCP

 Working Process:
 Advertising routes: Each AS shares the IP range to its neighbors.
 Path Selection: BGP selects the best path on the basis of path length,
speed and reliability, which may not always be the shortest path, but best
according to the policies.
 Routing Updates: Updates route only when its necessary. For example,
Router addition, removal, etc.